Impact of Mine Subsidence on Threatened Ecological Communities

The aim of this project was to develop and implement integrated remote sensing and field survey to quantify the impacts of longwall mine subsidence (LWMS) on Bluegrass and Brigalow, Threatened Ecological Communities (TEC). The methods and results are anticipated to permit broader extrapolation of this approach to current and future mines and contribute to evidence for offset requirements.

Three mine sites within the Bowen Basin were the focus of this study, being: Goonyella Riverside, Moranbah North and Kestrel.

The overall project design was structured to test the following hypotheses:

· Subsidence has a negative impact on native Brigalow communities in comparison to unsubsided areas (site scale);

· Differential impacts at Maximum Subsidence (LW), Transition (T) and Pillar (P) zones within each longwall will have differential impacts on Brigalow communities (within longwall panels);

· Impacts on Brigalow will be most noticeable over more recently subsided longwalls as rehabilitation time will be less (temporal scale);

· Subsidence has a negative impact on native Bluegrass communities in comparison to unsubsided areas (site scale);

· Differential impacts at LW, T and P zones within each longwall will have differential impacts on Bluegrass communities (within longwall panels);

· Impacts on Bluegrass will be most noticeable over more recently subsided longwalls as rehabilitation time will be less (temporal scale).

Analysis of both the remotely sensed (Lidar and imagery) and field data did not find any significant impacts on Brigalow community condition that were linked to LWMS. Surface land management (historical and present) have a much greater impact on the extent, distribution and condition of this TEC. The combination of Lidar and image derived parameters permitted the extraction and analysis of vegetation condition parameters including: woodland height, projected foliar cover, crown cover, patch size and relative leaf condition. While some parameters such as exotic species, large woody debris and hollows are difficult to assess with remotely sensed data the 'extractable' parameters combined with field assessment provide for detailed comparison of woodland condition. Also, remotely sensed data allows for assessment over large areas, providing a site or regional context and permitting comparison between impact and control.

At Kestrel, assessment of the Bluegrass communities did not show any impacts associated with LWMS. Using the historical image set and earlier field survey data, obvious surface management practices were described.  It was found that preferential management of the Bluegrass pastures had allowed a substantial recovery of these communities with a concomitant reduction in introduced non-native pasture species and weeds.  As with Goonyella Riverside and Moranbah North, surface management was found to have a much greater impact on the condition of these communities than subsidence.

Each of the hypotheses were rejected as there was no clear evidence from visual assessment or statistical comparison that LWMS reduced the condition of these communities.  Any impacts that may have occurred on these communities due to subsidence were either small or obscured by vegetation regeneration.

While this study showed no consistent difference between these zones, stratifying sampling into predicted impact zones is still recommended.  Further, site specific details such as: seam depth; angle of draw; seam extraction thickness; surface slope; soil type and land use should continue to be considered for future studies.